I haven't checked Sandra's link (the Grauniad web site gives my browser indigestion), but Sir Arthur Evans also did damage to the understanding of Linear B, by insisting it was a native Cretan language; such was his authority that nobody dare contradict him. Only in the 1940s, after Evans was long dead, did Alice Kober get hold of samples and start cracking Linear B from scratch; Michael Ventris then found it to be Greek after all.

Reference: Simon Singh: The Code Book, ch 5 ("The language barrier"). Read the whole saga (pp 217--242) for the connection with Bletchley Park.